In this paper, we present a multivariate neural network model as an innovative method for stock value prediction. For this method, we con-sider 10 statistical values given by the financial time series data: the stock it-self, an asset used as a reference, RSI, Trix, Fisher, MACD, Mayer regression, Mayer and correlation. Datasets are based on daily time intervals taken from YFinance API. The model architecture comprises multiple dense layers with dropout regularization to mitigate overfitting. The approach we have proposed is effective for forecasting stock price trends in the daily stock market, which we observed in the experimental results. This shows a re-markable accuracy on a test dataset. Our research aids to the advancement of predictive modeling techniques in financial market.